The Importance of Matchmaking in League of Legends 1 Jon Decelle - - PowerPoint PPT Presentation

the importance of matchmaking in league of legends
SMART_READER_LITE
LIVE PREVIEW

The Importance of Matchmaking in League of Legends 1 Jon Decelle - - PowerPoint PPT Presentation

The Importance of Matchmaking in League of Legends 1 Jon Decelle Gabriel Hall Lindsay ODonnell Section 1 2 Introduction and Problem Statement Online Games 3 In online games, players connect to one another over a network and play


slide-1
SLIDE 1

The Importance of Matchmaking in League of Legends

Jon Decelle Gabriel Hall Lindsay O’Donnell

1

slide-2
SLIDE 2

Introduction and Problem Statement

Section 1

2

slide-3
SLIDE 3

Online Games

  • In online games, players connect to one another over a network and play with

each other.

  • Some games are 1v1, while others are team based.
  • In many of these games, players’ skill levels can vary widely.
  • For enjoyment, players want to play with other players of similar skill.

3

slide-4
SLIDE 4

Matchmaking

  • Matchmaking places players of similar skill together in a

match.

  • Better matchmaking algorithms form more even matches.
  • Matchmaking needs to be fast

4

slide-5
SLIDE 5

Multiplayer Online Battle Arena (MOBA)

  • Competitive, team based
  • Win the game through

killing enemy structures

  • MOBAs are session based,

so players develop skill and increase their ranking by winning games

5

slide-6
SLIDE 6

League of Legends (LoL)

  • Most popular current day MOBA [Forbes 2012]
  • Free to Play
  • Developed by Riot Games in 2009
  • WPI has many LoL players to study

6

slide-7
SLIDE 7

Ranking and Matchmaking in LoL

  • Each player has a visible rank and a hidden matchmaking rating

(MMR)

  • This rating is based on the Elo formula, originally designed for

chess

  • LoL ranks: Bronze, Silver, Gold, Platinum, Diamond, Master,

and Challenger.

7

slide-8
SLIDE 8

Research Questions

  • How balanced is matchmaking in LoL?
  • How long do players wait for a match?
  • How do players perceive how well matchmaking works?
  • What factors affect player enjoyment?

8

slide-9
SLIDE 9

Study Methodology

Section 2

9

slide-10
SLIDE 10

Study Methodology Overview

10

  • Collect player ranks to determine how evenly matched

teams were by rank

  • Collect player perceptions of how even game turned out
  • Collect game data to show how even game was statistically
  • Compare these to see how much of an impact matchmaking

has on game

slide-11
SLIDE 11

Study Advertising

11

  • Solicited subjects by emailing IMGD majors, Gaming club,

and asking friends

  • Offered open time slots for subjects to come in
  • Held studies on the weekend when more people are

generally free

slide-12
SLIDE 12

Study Regulations

12

  • Players signed consent forms
  • Players set up their game settings
  • Asked to play 5v5 games, solo queue only as this affects the

matchmaking

  • Play Ranked if possible

○ Uses their Ranked MMR which is represented by their

visible rank

slide-13
SLIDE 13

Surveys

13

  • Collected demographic player information pre-game
  • Collected player perceptions post-game
  • Done before and after each game

○ Game was fresh in player’s mind

slide-14
SLIDE 14

Environment

14

  • Computer lab where all the players could see, hear,

and talk to one another during games.

  • Many subjects came in groups

○ Although they were playing separately, often

interacted

slide-15
SLIDE 15

Matchmaking Data Collection

15

  • Collected ranks of each player in each game through

third party website while game in progress

  • lolnexus.com
slide-16
SLIDE 16

Player Survey

16

  • After each game:

○ Game enjoyability ○ Game evenness overall and at different times during game ○ Reasons for win or loss ○ Outlier players

slide-17
SLIDE 17

Game Data

17

  • Collected game data from Riot’s match history website:

○ Gold difference at 20 minutes and at the end of game ○ Kills for each team ○ Game time ○ KDA for each player ○ Position played

slide-18
SLIDE 18

Even vs Uneven

  • There are many measures to show how even or uneven a

game was:

○ Kill difference ○ Gold difference (Gold is obtained by killing enemies and

enemy structures)

○ Game Time (If one team is far ahead, they are likely to finish

the game quickly)

18

slide-19
SLIDE 19

Analysis and Results

Section 3

19

slide-20
SLIDE 20

Study Results and Demographics

  • 23 unique participants
  • 3 studies, 5 hours each
  • 65 games, 52 complete and valid surveys

○ Website for data collection went down during one study

20

slide-21
SLIDE 21

How does queue time correlate with rank?

21

slide-22
SLIDE 22

Does better teamwork lead to higher player enjoyment?

  • Axes:

○ x: “How well did your team work together?” ○ y: “How enjoyable was was the game you just played?”

  • The line of fit shows the

positive correlation between these responses.

22

slide-23
SLIDE 23

Are even games more fun?

  • Axes:

○ x: “How even was the game you just played?” ○ y: “How enjoyable was was the game you just played?”

23